Back in my RACAL days, manufacturing made a point of hand fitting small glass beads on the leads of the bead tantalum capacitors, just to reduce cracking the epoxy. There must have been a reason for this expensive process.
Nothing strikes me as at all obvious, having made many thousands when I worked for AVX tantalum.
The encapsulant is just a "wrapper" to hold the assembly together - no doubt there will be an X-ray of one onlne somewhere, so that you can see most of the assembly.
If there is any damage to the actual capacitive element (which is essentially the anode - the sintered and anodised tantalum "sponge"), it would most likely go SC, with a far smaller chance of going OC.
To give you an idea of how robust the things are, the way that we checked % coverage of the dielectric (tantalum pentoxide), with the cathode material (manganese dioxide), a sample of cap's were cooked in a pressure cooker for an hour or so, so that water became the cathode material, filling the "sponge" and pretty much covering all of the dielectric. Sure, you would get a few total duds, but most survived to be able to measure them. They were measured before cooking, so you had a measure of how much dielectric was covered with cathode material.
In recent years, minute PTFE washers were used by AVX, on the Ta leadwire that formed part of the anode. This was before anodising and was to stop/restrict cathode material migrating up the leadwire, which would give a short as the leadwire wasn't anodised. Everything was mechanised and the assembly of the washers to the leadwires and the anodes assembled to what were/are called stringers - strips of SS or aluminium that held many anodes, like washing on a line - had already moved to the Czech Republic before I worked there, so I never saw it.
